Adwiteey Mauriya, Ph.D.
Senior Data Science Engineer | ML Engineering & Gen AI | LLM | Fintech
I am a Senior Data Science Engineer, former Physicist creating data pipeline, maintaining data quality and governance and building/training effective machine learning models on various projects.
About Me
I have 10 years of experience in machine learning and 7 years in big data engineering, 4 years of experience in LLM, specialising in scalable data pipelines, automation, and AI-driven insights. I’ve worked with Snowflake Cloud, Apache Airflow, and Python to process large datasets efficiently. Recently, I’ve focused on generative AI and LLMs in fintech, optimizing financial analytics and risk assessment.
Professional Summary
As the Founder of Flaredog, I lead innovative AI/ML infrastructure solutions that transform industries such as finance, healthcare, climate, and blockchain. With over ten years of research and industry experience, I combine my expertise in AI, data science, and engineering to help businesses tackle complex challenges and build scalable, high-performance solutions. Prior to founding Flaredog, I applied my skills in fintech, delivering impactful AI solutions, including large-scale machine learning models and data pipelines. I hold a Ph.D. in Computational Plasma Physics from Instituto Superior Tecnico, Lisbon, Portugal, and have a strong background in statistical data analysis, high-performance computing, and data analytics. I am passionate about driving technological innovation and empowering businesses through data and AI.
Key Achievements
- •UAE Golden visa holder in Talented Pioneers Category
- •Top 4% in most difficult engineering exam in India (IIT-JEE) 2009
- •Received competitive KVPY fellowship in 2009 (for 5 years), funded by Dept. of Science and Tech and Government of India
- •Received prestigious FCT fellowship and Eurofusion research grants
- •Obtained All India Rank 185 in India's National Ph.D. Qualifying Exam (CSIR NET)
Contact Information
Permanent Address
5°, 6E, Largo Jorge Moutinho Albuquerque, Lisbon, 2780-255 PT
Phone
+351 927 728 594
+44 20 3286 2669
+971 58 569 1933
Languages
Work Experience
My professional journey across various roles and organizations.
Founder (Self Employed)
Flaredog | Zurich, Switzerland
- •Launched Flaredog, a business solutions company specializing in AI/ML infrastructure.
- •Developed and deployed 5+ AI/ML solutions across finance, healthcare, climate, and blockchain, leading to an average 30% efficiency boost in client workflows.
- •Led business development initiatives, client acquisition, and cross-functional project delivery.
- •Implemented cost-efficient ML infrastructure, reducing cloud computing expenses by 25% while improving model inference speed by 50%.
Senior Data Science Engineer
Liquidity Capital Inc. | Abu Dhabi, UAE
- •Developed and deployed a financial chatbot powered by LLMs & LangChain, handling 5,000+ document-based queries monthly, reducing manual financial analysis time by 40%.
- •Optimised probability of default (PD) model, increasing prediction accuracy by 15%, enabling data-driven risk assessments and enhancing underwriting efficiency by 30%.
- •Built a centralised feature store, consolidating 100+ financial indicators from structured & unstructured data, improving model retraining efficiency by 50%.
- •Automated active vs. out-of-business classification, increasing classification accuracy by 92%, reducing manual verification efforts by 60%.
- •Deployed a BI tool on EKS, enabling real-time insights for 50+ stakeholders, reducing report generation time by 35% and enhancing decision-making efficiency.
- •Enhanced model inference speeds by 3X through optimization in Databricks & Torch, cutting down compute costs by 25%.
- •Collaborated with cross-functional teams, integrating AI-driven insights that led to a 20% improvement in investment decision accuracy.
Senior Data Engineer
Liquidity Capital Inc. | Abu Dhabi, UAE
- •Led the development of a bank account aggregator pipeline, processing 2M+ transactions monthly, improving data ingestion efficiency by 90%.
- •Automated ETL workflows, reducing data processing time from 16 hours to 10 minutes, enhancing reporting speeds by 80%.
- •Implemented an ML-driven anomaly detection system, cutting down fraudulent transaction alerts by 50%, increasing accuracy in risk assessment.
- •Designed a CI/CD pipeline for ETL repositories, reducing deployment time by 60% and minimising errors in production releases.
- •Managed AWS EC2 instances & AWS Lambda functions, reducing infrastructure costs by 30% through auto-scaling optimizations.
- •Developed and maintained Delta Lake architecture, enabling 99.9% data availability across financial datasets.
- •Optimized Spark-based transformations, achieving 2.5X faster query execution, reducing cloud computing costs by 20%.
- •Established backend code review best practices, improving deployment efficiency and reducing rollback incidents by 60%.
Data Engineer
Tripadvisor Inc. | Lisbon, Portugal
- •Maintained and monitored datasets across 3+ major cloud platforms (GCP, Snowflake, Hadoop), ensuring 99.98% data availability for analytics and reporting.
- •Optimised ETL pipelines, reducing data processing time by 50%, which improved dashboard refresh speeds for business intelligence teams.
- •Implemented data compression techniques, lowering storage costs by 40% and contributing to $10M+ annual cost savings.
- •Designed and built 30+ new data pipelines, automating ingestion and transformation for 200M+ records per day.
- •Improved query performance by 3x in Snowflake and BigQuery, enabling near real-time analytics for marketing and product teams.
- •Led the migration of legacy ETL workflows to Airflow, reducing manual interventions by 80%, increasing job success rates to 99.5%.
- •Enhanced data quality checks, reducing data discrepancies by 90%, improving reporting accuracy for finance and customer teams.
- •Championed best practices in code reviews, leading to a 50% reduction in deployment errors and improved team efficiency.
- •Trained and mentored 2+ junior data engineers, accelerating their onboarding process and reducing ramp-up time by 30%.
Business Data Analyst
Netjets Inc. | Oeiras, Portugal
- •Automated pilot training data warehouse, reducing manual data entry workload by 70%, improving accuracy, and increasing operational efficiency.
- •Developed 20+ dynamic reports and dashboards using Tableau, Power BI, and SSRS, enabling executives to make data-driven decisions in real time.
- •Increased data processing speed by 3x, improving insights delivery for training performance analysis.
- •Designed KPI tracking framework, helping the company optimise pilot training costs by 15% while maintaining compliance with aviation standards.
- •Achieved 110% of the targeted qualified leads from strategic training workshops, exceeding conversion expectations.
- •Implemented data validation workflows, reducing data inconsistencies by 90%, ensuring high-quality analytics.
- •Improved gross margin tracking, leading to a 5% increase in profitability for the training department by identifying inefficiencies.
- •Optimized SQL queries and ETL pipelines, reducing report generation time by 50%, enabling faster decision-making for leadership.
Researcher
University of Oxford, Culham Center for Fusion Energy | Oxfordshire, UK
- •Developed and optimized numerical tools to simulate plasma behavior in nuclear fusion reactors, improving prediction accuracy by 35% and reducing computational time by 40% on high-performance clusters.
- •Scaled large-scale simulations on supercomputers with 100,000+ CPU cores and GPU nodes, enhancing processing efficiency 3x while reducing energy consumption by 20% through optimized resource allocation.
- •Designed an automated pipeline for importing and analyzing fusion experiment data, reducing manual processing time by 60% and improving data throughput by 50% using advanced compression techniques.
- •Reduced computational errors by 30% by implementing improved numerical solvers and stability corrections in simulation codes.
Project Scientist
Institute for Plasma Research | Gandhinagar, India
- •Created a code in python and Fortran 90 to with more than 100% better time complexity with the existing code.
- •Implemented the algorithm in a parallel HPC code.
- •Merged the new code with existing fix boundary equilibrium code.
- •Patched the merged code into a Fortran library and integrated it with a multiphysics code.
Linux Sysadmin
Soft Matter Simulaton Laboratory, IISER | Pune, India
- •Provided technical support to entire research group to onboard and updating their machines which includes security updates, patching and compliance best practices
- •Assisted in setting up scientific software and libraries and help them compiling and debugging them by setting up necessary tools
- •Debugged High-Performance Computing codes e.g Gromacs.
- •Automated various tasks through coded batch jobs in Bash and Python.
Founder (Self Employed)
Flaredog | Zurich, Switzerland
- •Launched Flaredog, a business solutions company specializing in AI/ML infrastructure.
- •Developed and deployed 5+ AI/ML solutions across finance, healthcare, climate, and blockchain, leading to an average 30% efficiency boost in client workflows.
- •Led business development initiatives, client acquisition, and cross-functional project delivery.
- •Implemented cost-efficient ML infrastructure, reducing cloud computing expenses by 25% while improving model inference speed by 50%.
Senior Data Science Engineer
Liquidity Capital Inc. | Abu Dhabi, UAE
- •Developed and deployed a financial chatbot powered by LLMs & LangChain, handling 5,000+ document-based queries monthly, reducing manual financial analysis time by 40%.
- •Optimised probability of default (PD) model, increasing prediction accuracy by 15%, enabling data-driven risk assessments and enhancing underwriting efficiency by 30%.
- •Built a centralised feature store, consolidating 100+ financial indicators from structured & unstructured data, improving model retraining efficiency by 50%.
- •Automated active vs. out-of-business classification, increasing classification accuracy by 92%, reducing manual verification efforts by 60%.
- •Deployed a BI tool on EKS, enabling real-time insights for 50+ stakeholders, reducing report generation time by 35% and enhancing decision-making efficiency.
- •Enhanced model inference speeds by 3X through optimization in Databricks & Torch, cutting down compute costs by 25%.
- •Collaborated with cross-functional teams, integrating AI-driven insights that led to a 20% improvement in investment decision accuracy.
Senior Data Engineer
Liquidity Capital Inc. | Abu Dhabi, UAE
- •Led the development of a bank account aggregator pipeline, processing 2M+ transactions monthly, improving data ingestion efficiency by 90%.
- •Automated ETL workflows, reducing data processing time from 16 hours to 10 minutes, enhancing reporting speeds by 80%.
- •Implemented an ML-driven anomaly detection system, cutting down fraudulent transaction alerts by 50%, increasing accuracy in risk assessment.
- •Designed a CI/CD pipeline for ETL repositories, reducing deployment time by 60% and minimising errors in production releases.
- •Managed AWS EC2 instances & AWS Lambda functions, reducing infrastructure costs by 30% through auto-scaling optimizations.
- •Developed and maintained Delta Lake architecture, enabling 99.9% data availability across financial datasets.
- •Optimized Spark-based transformations, achieving 2.5X faster query execution, reducing cloud computing costs by 20%.
- •Established backend code review best practices, improving deployment efficiency and reducing rollback incidents by 60%.
Data Engineer
Tripadvisor Inc. | Lisbon, Portugal
- •Maintained and monitored datasets across 3+ major cloud platforms (GCP, Snowflake, Hadoop), ensuring 99.98% data availability for analytics and reporting.
- •Optimised ETL pipelines, reducing data processing time by 50%, which improved dashboard refresh speeds for business intelligence teams.
- •Implemented data compression techniques, lowering storage costs by 40% and contributing to $10M+ annual cost savings.
- •Designed and built 30+ new data pipelines, automating ingestion and transformation for 200M+ records per day.
- •Improved query performance by 3x in Snowflake and BigQuery, enabling near real-time analytics for marketing and product teams.
- •Led the migration of legacy ETL workflows to Airflow, reducing manual interventions by 80%, increasing job success rates to 99.5%.
- •Enhanced data quality checks, reducing data discrepancies by 90%, improving reporting accuracy for finance and customer teams.
- •Championed best practices in code reviews, leading to a 50% reduction in deployment errors and improved team efficiency.
- •Trained and mentored 2+ junior data engineers, accelerating their onboarding process and reducing ramp-up time by 30%.
Business Data Analyst
Netjets Inc. | Oeiras, Portugal
- •Automated pilot training data warehouse, reducing manual data entry workload by 70%, improving accuracy, and increasing operational efficiency.
- •Developed 20+ dynamic reports and dashboards using Tableau, Power BI, and SSRS, enabling executives to make data-driven decisions in real time.
- •Increased data processing speed by 3x, improving insights delivery for training performance analysis.
- •Designed KPI tracking framework, helping the company optimise pilot training costs by 15% while maintaining compliance with aviation standards.
- •Achieved 110% of the targeted qualified leads from strategic training workshops, exceeding conversion expectations.
- •Implemented data validation workflows, reducing data inconsistencies by 90%, ensuring high-quality analytics.
- •Improved gross margin tracking, leading to a 5% increase in profitability for the training department by identifying inefficiencies.
- •Optimized SQL queries and ETL pipelines, reducing report generation time by 50%, enabling faster decision-making for leadership.
Researcher
University of Oxford, Culham Center for Fusion Energy | Oxfordshire, UK
- •Developed and optimized numerical tools to simulate plasma behavior in nuclear fusion reactors, improving prediction accuracy by 35% and reducing computational time by 40% on high-performance clusters.
- •Scaled large-scale simulations on supercomputers with 100,000+ CPU cores and GPU nodes, enhancing processing efficiency 3x while reducing energy consumption by 20% through optimized resource allocation.
- •Designed an automated pipeline for importing and analyzing fusion experiment data, reducing manual processing time by 60% and improving data throughput by 50% using advanced compression techniques.
- •Reduced computational errors by 30% by implementing improved numerical solvers and stability corrections in simulation codes.
Project Scientist
Institute for Plasma Research | Gandhinagar, India
- •Created a code in python and Fortran 90 to with more than 100% better time complexity with the existing code.
- •Implemented the algorithm in a parallel HPC code.
- •Merged the new code with existing fix boundary equilibrium code.
- •Patched the merged code into a Fortran library and integrated it with a multiphysics code.
Linux Sysadmin
Soft Matter Simulaton Laboratory, IISER | Pune, India
- •Provided technical support to entire research group to onboard and updating their machines which includes security updates, patching and compliance best practices
- •Assisted in setting up scientific software and libraries and help them compiling and debugging them by setting up necessary tools
- •Debugged High-Performance Computing codes e.g Gromacs.
- •Automated various tasks through coded batch jobs in Bash and Python.
Technical Skills
A comprehensive overview of my technical expertise and capabilities.
Big Data
ETL
Database
DevOps
Machine Learning
BI
Development
Cloud
Education
My academic background and qualifications.
Formal Education
Internships
Certifications & Fellowships
Professional certifications and academic fellowships
Professional Certifications
Fellowships & Grants
Ph.D. Scholarship
Fundação para a Ciência e Tecnologia, Lisbon, Portugal
KVPY Fellowship
Department of Science and Technology, India
Junior Research Fellowship, All India rank 185
Council for Scientific and Industrial Research (CSIR), Government of India
Eurofusion allowance
Eurofusion consortium, Garching, Germany
Collaborator of MARCONI-Eurofusion supercomputing project
CINECA Super computing application and Innovation department, Bologna, Italy
Testimonials
What colleagues and clients say about working with me
Claudia Ceboleiro
Strategic Initiatives PMO & Continuous Improvement Project Manager at NetJets Europe
"Adwiteey have accepted to make part of a global Training Management System implementation project I manage, as a report builder, required by project to support system functionalities. Despite his lack of business aviation knowledge, very quickly Adwiteey has adapted to our European and American work environments and Project Subject Matter Experts. Technically Adwiteey has demonstrated his enormous knowledge of Structured Query Language(SQL) and was able to develop complex queries and manipulate databases. His phenomenal analytical and reasoning skills were not limited to build reports but also to add value to the product by suggesting innovative solutions, solving problems as they arise, understanding at the same time system and business limitations and supporting the project team as required. He was a very useful piece of our puzzle we will lose now due to budget limitations. He has a very special spark and I trust he will go far by making our world a better place."
Get in Touch
Have a new job position in your organisation? Let's chat and find out. We can make a difference together.
Send a Message
Contact Information
Phone
+351 927 728 594
+44 20 3286 2669
+971 58 569 1933
Location
Zurich, Switzerland